Spaces:
Running
on
CPU Upgrade
trying to create a space with the open model, i am new please help.
Traceback (most recent call last):
File "app.py", line 13, in
pipe = StableDiffusionPipeline.from_pretrained(model_id, use_auth_token=True, revision="fp16", torch_dtype=torch.float16)
File "/home/user/.local/lib/python3.8/site-packages/diffusers/pipeline_utils.py", line 154, in from_pretrained
cached_folder = snapshot_download(
File "/home/user/.local/lib/python3.8/site-packages/huggingface_hub/utils/_deprecation.py", line 93, in inner_f
return f(*args, **kwargs)
File "/home/user/.local/lib/python3.8/site-packages/huggingface_hub/_snapshot_download.py", line 115, in snapshot_download
raise EnvironmentError(
OSError: You specified use_auth_token=True, but a Hugging Face token was not found.
this is a demo not a model
how would one make a demo like this ?
Try the colab version
Hi
@iceborg
, in order to make a demo on a Spaces or for local deployment, you would need to create a User Access Token here https://huggingface.co/settings/tokens and add this token to the Repo secrets here:
Then visit https://huggingface.co/CompVis/stable-diffusion-v1-4, read and accept the LICENSE if you agree
However, this code won't work out of the box on cloned Spaces because right now Spaces that use GPU are restricted to some demos. In order to run on CPU you also have to not use the autocast part of the code. I have updated the codebase with some comments to make this process easier to follow
In case you are running locally, you can either:
- Run
huggingface-cli loginor - Paste the token here
pipe = StableDiffusionPipeline.from_pretrained(model_id, use_auth_token="YOUR_GENERATED_TOKEN", revision="fp16", torch_dtype=torch.float16)
@iceborg also to add to this, to make a gradio demo you can follow this google colab notebook: https://colab.research.google.com/drive/1NfgqublyT_MWtR5CsmrgmdnkWiijF3P3?usp=sharing